Community structure detection for the functional connectivity networks of the brain

Abstract

The community structure detection problem in weighted networks is tackled with a new approach based on game theory and extremal optimization, called Weighted Nash Extremal Optimization. This method approximates the Nash equilibria of a game in which nodes, as players, chose their community by maximizing their payoffs. After performing numerical experiments on synthetic networks, the new method is used to analyze functional connectivity networks of the brain in order to reveal possible connections between different brain regions. Results show that the proposed approach may be used to find biomedically relevant knowledge about brain functionality.